| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 NYU School of Medicine, New York, NY
2 Max Plank Institute, Heidelberg, Germany
* Corresponding author: egilland{at}mbl.edu
Both luminescent (1, 2) and fluorescent (2) reporters have been used to image periodic large-scale intercellular calcium waves that begin during zebrafish gastrulation, at about 65% epiboly, and continue for at least 12 hours. These waves arise every 5 to 10 min from a variety of locations and traverse the blastoderm margin and main body axis (1). During somitogenesis they appear as a series of pulses of elevated calcium levels centered on the tailbud region. The waves travel at approximately 5–10 µm per s and thus fall into the category of fast calcium waves that likely propagate by a positive feedback mechanism involving calcium-induced calcium release from intracellular stores, possibly including diffusion of calcium or IP3 through gap junctions (3). Likely targets of the waves include calcium-sensitive proteins involved in epiboly (3) and convergent extension (4), and others such as calreticulin that may play a role in the temporal regulation of nuclear receptor activity (5). Long-distance signaling by rhythmic calcium waves is an appealing mechanism for synchronizing calcium-triggered events throughout the embryo with high temporal precision. Since the zebrafish embryo is a roughly spherical body approximately 600 µm in diameter, imaging these waves in a single optical plane, as in previous studies, can only approximate their three-dimensional trajectories. Moreover, other calcium signals that may be occurring at the same time, but in different optical planes within the embryo, cannot be documented. Ideally, free calcium levels should be imaged for many hours throughout the entire volume of the embryo at intervals shorter than the lifetimes of individual signaling events. Luminescent techniques require considerable temporal integration to achieve adequate spatial resolution and thus cannot approach this goal. Conversely, scanning laser microscopy using visible light has the necessary spatial and temporal resolution, but cannot be used for prolonged imaging at high sampling rates due to phototoxic damage to the embryo (E. Gilland, pers. obs.). The present study demonstrates that multiphoton fluorescence microscopy has the potential to achieve the goal of sampling calcium dynamics throughout the entire zebrafish embryo for long durations with sufficient spatial and temporal resolution to reveal complex three-dimensional signaling events.
Glass micropipettes with outside tip diameters of 7 to 10 µm were used to pierce the chorion and inject zebrafish embryos (2- to 4-cell stages) with 1 to 2 nl of a 2% solution of Fluo-4 dextran (calcium Kd = 3 µM; MW = 10 kDa; Molecular Probes) dissolved in 0.2 M KCl. The injections were made by a vegetal approach and placed into the dorsal half of the yolk mass. Dye was carried into the cytoplasm by diffusion and cytoplasmic streaming. Embryos were injected while in 30% Danieau’s solution (1) and then transferred to 10% Danieau’s solution, placed into 1 mm diameter grooves in agarose-coated dishes, and maintained at 28 °C on a thermoelectric heating stage. Imaging was performed with a custom-built multiphoton fluorescence microscope using 890 nm excitation light from a 5-W Mira titanium-sapphire pulsed laser (Coherent) and a 20x, wide aperture, infrared objective, N.A. 0.8 (Olympus). Embryos were imaged in a series of optical planes, either parasagittal, frontal, or horizontal, starting just below the surface of the embryo, with successive planes separated along the z axis by distances from 5 to 20 µm, for total imaging depths of up to approximately 300 µm. Scanning each image plane in a 256 x 256 pixel pattern required 0.4 to 0.8 s, depending on the beam dwell time. In a typical experiment, a sequence of 15 optical slices was repeated every 14 s over 8 h. Collected image sequences were analyzed using the IDL (SRI) and ImageJ (NIH) software packages.
The experiments showed that the periodic increases in calcium levels seen in previous studies (1, 2) pass through all cellular regions of the embryo and, at later stages, through most or all of the yolk cell. Figure 1 shows data from an experiment in which 45 calcium waves were imaged over 4 h. Time lapse movies of the imaging sequences can be viewed on the Biological Bulletin’s web site (www.mbl.edu/BiologicalBulletin/VIDEO/BB.video.html). Single images (X-Y) from different parasagittal optical planes during one imaging cycle (15 planes in 14.3 s) are shown in Figure 1, a–e. A plot of pixel values through time, for an area dorsal to the tailbud, shows calcium levels oscillating with a mean period of 315 s over many hours (Fig. 1f). As found in previous studies (1, 2), the strongest modulation of calcium levels was seen in the caudal half of the embryo centered around the region of the tail bud, with the brightest signals in the mesoderm, endoderm, and subjacent yolk syncytial layer. Since cellular densities and dye distribution (not shown) vary in different tissue layers, ratio imaging will be required to determine whether the greater intensities of signal in specific locations represent genuinely higher calcium levels. The movement of calcium waves through the embryo can be roughly assessed by eye in the time lapse movies, and more precisely, either by comparing the time coordinates of peaks at different X-Y-Z locations (not shown), or by reslicing an X-Y time series along a line in the X-Y plane (Fig. 1g). The calcium oscillations appear as light and dark bands in the resliced image, and the slopes of the bands indicate the direction of wave movement. In the case shown, most of the oscillations appeared earliest at dorsal and ventral locations, and latest at a region dorso-anterior to the tail bud, indicating that the waves were sweeping caudally through the embryo towards the tail bud. Other waves showed the inverse pattern, i.e., they started near the tail bud and traveled rostrally. The three-dimensional structure of the calcium waves hinted at by these initial experiments are far more complex than could be inferred from sequences imaged at single optical planes. To completely reconstruct the waves through space and time, the temporal phase of wave peak values must be quantified for all X-Y positions in all optical planes through the time series. Such analyses may provide further clues to the developmental role of the panembryonic calcium signaling system.
|
Literature Cited
This article has been cited by other articles:
|
|
 |
|
  M. Whitaker Calcium at Fertilization and in Early Development Physiol Rev, January 1, 2006; 86(1): 25 - 88. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
